The central goals of this competing renewal grant are to: (1) identify specific mediated pathways of risk for substance use and abuse among sexual minority girls (SMGs; girls who report same-sex sexual orientation) and (2) identify developmentally appropriate protective factors that can inform the development of novel prevention and intervention studies tailored for this high risk group. SMGs are 400% more likely to report substance use than are heterosexual girls. Recent results from grant years 1-3 showed that SMGs reported multiple mental health disparities including risky sexual behavior that were significant across all levels of all demographic covariates. SMG substance use disparities seen in adolescence increased as girls transition to young adulthood and change in binge drinking over time is more closely linked to suicidality in SMGs than it is in boys. In grant years 1-3 we were highly successful in documenting the persistence of SMG substance use disparities and identifying important psychosocial correlates of the disparities. Building on this foundation, we propose a heuristic model that articulates the most probable (empirically-supported, theory-driven) pathways of risk for substance use in SMGs. Dozens of conceptual models have been developed to describe risk for substance use among other at-risk teens; however no one has developed an empirically supported heuristic model that describes risk pathways for sexual minority youth. Preliminary data from grant years 1-3 and our proposed data sets (Add Health and the Pittsburgh Girls Study) show strong preliminary support for our hypothesized mediate pathways (stress-negative affect, social marginalization, and future expectations pathways). This proposal will build upon this foundation, and seeks to move the field forward by proposing and testing several longitudinal mediated pathways of risk and identify modifiable protective factors that will inform the design of a novel prevention and intervention program for SMGs.